EBI is an Outstation of the European Molecular Biology Laboratory.

Transcription

1 EBI is an Outstation of the European Molecular Biology Laboratory.

2 InterPro is a database that groups predictive protein signatures together 11 member databases single searchable resource provides functional analysis of proteins by classifying them into families and predicting domains and important sites Enables whole genome analysis

3 InterPro Consortium Consortium of 11 major signature databases

4 Protein signatures More sensitive homology searches Each member database creates signatures using different methods and methodologies: manually-created sequence alignments automatic processes with some human input and correction entirely automatically.

5 Why do we need predictive annotation tools?

6 What are protein signatures? Protein family/domain Multiple sequence alignment Build model Search Protein analysis it. UniProt Significant match ITWKGPVCGLDGKTYRNECALL AVPRSPVCGSDDVTYANECELK Mature model

7 Member databases METHODS Hidden Markov Models Finger- Prints Profiles Patterns Sequence Clusters Structural Domains Protein features (active sites ) Functional annotation of families/domains Prediction of conserved domains

8 InterPro entry

9 InterPro entry

10 The InterPro entry: types Family Proteins share a common evolutionary origin, as reflected in their related functions, sequences or structure Domain Distinct functional, structural or sequence units that may exist in a variety of biological contexts Repeats Short sequences typically repeated within a protein Sites PTM Active Site Binding Site Conserved Site

11 InterPro Entry Groups similar signatures together Adds Adds extensive extensive annotation Links Links to other to other databases databases Structural information and viewers Quality control Removes redundancy

12 InterPro Entry Groups similar signatures together Adds Adds extensive extensive annotation Links Links to other to other databases databases Structural information and viewers Hierarchical classification

13 Interpro hierarchies: Families FAMILIES can have parent/child relationships with other Families Parent/Child relationships are based on: Comparison of protein hits child should be a subset of parent siblings should not have matches in common Existing hierarchies in member databases Biological knowledge of curators

14 InterPro hierarchies: Domains DOMAINS can have parent/child relationships with other domains

15 Domains and Families may be linked through Domain Organisation Hierarchy

16 InterPro Entry Groups similar signatures together Adds Adds extensive extensive annotation Links to other databases Links to other databases Structural information and viewers

17 InterPro Entry Groups similar signatures together Adds extensive annotation Adds extensive annotation Links Links to other to other databases databases Structural information and viewers The Gene Ontology project provides a controlled vocabulary of terms for describing gene product characteristics

18 InterPro Entry Groups similar signatures together Adds extensive annotation Adds extensive annotation Links Links to other to other databases databases Structural information and viewers UniProt KEGG... Reactome... IntAct... UniProt taxonomy PANDIT... MEROPS... Pfam clans... Pubmed

19 InterPro Entry Groups similar signatures together Adds extensive annotation Adds extensive annotation Links to other databases Links to other databases Structural information and viewers PDB 3-D Structures SCOP Structural domains CATH Structural domain classification

20 InterProScan Protein Sequence Analysis algorithm Raw Matches Filtering algorithm Reported Matches Predictive Models

21 InterProScan access Interactive: Webservice (SOAP and REST): Downloadable: ftp://ftp.ebi.ac.uk/pub/software/unix/iprscan/

22 Why redesign InterProScan? InterProScan 4 complicated installation complicated update limited queuing system Only guaranteed with LSF limited configurability reliability

23 InterProScan 5.0 aims Easy install and configuration Modular Expandable Easily integrated into existing pipelines Incorporate new data model / XML exchange format Easy to port on to different architectures: Desktop machine Simple LAN LSF PBS Sun Grid Engine...cloud? GRID? Reliablity

24 InterProScan 5 Technology

25 Architecture Cluster platform Web services Java API InterPro website JMS: monitoring queues Job Management: scheduling analyses Business Logic: performing analyses One-way dependencies + replaceable layers = low-coupling + maintainable Oracle PostgreSQL HSQLDB Database Access XML Data Model File I/O File system

26 Java Messaging Service Master Schedules tasks & sub-tasks, and places on queue Broker Manages queues & topics Broker starts workers on demand Workers take tasks off queues Monitoring & Management Application Web or stand-alone app to monitor & manage InterProScan Worker Peforms Worker task / sub-task Peforms Worker and task / reports sub-task Peforms Worker back Worker and to task / Worker reports Broker sub-task Peforms Peforms back and to task / task / reports Broker sub-task back and Performs to sub-task and task / reports Broker back to sub-task, reports Broker back reports to back Broker to Broker Simple and robust programming model Mature and stable standard current JMS version released in 2002 Guaranteed message delivery to a single worker Easy to monitor Flexible easy to implement on multiple platforms

27 Beta release functionality

28 Installation Requirements Java 1.6 Linux Perl Installation process wget ftp://ftp.ebi.ac.uk/pub/software/unix/iprscan/i5 dist.tar.gz tar xzf i5 dist.tar.gz ready to use

29 ./interproscan.sh i test_proteins.fasta o test_proteins.tsv goterms A2YIW7 f927b0d241297dcc9a1c5990b58bf3c4 122 Pfam PF00085 Thioredoxin E 28 T IPR Thioredoxin domain Biological Process:cell redox homeostasis (GO: ) A2YIW7 f927b0d241297dcc9a1c5990b58bf3c4 122 ProSitePatterns PS00194 Thioredoxin family active site T IPR Thioredoxin, conserved site Biological Process:cell redox homeostasis (GO: ) A2YIW7 f927b0d241297dcc9a1c5990b58bf3c4 122 PIRSF PIRSF null E 27 T IPR Thioredoxin Molecular Function:protein disulfide oxidoreductase activity (GO: ), Biological Process:glycerol ether metabolic process (GO: ), Biological Process:cell redox homeostasis (GO: ), Molecular Function:electron carrier activity (GO: ) A2YIW7 f927b0d241297dcc9a1c5990b58bf3c4 122 PRINTS PR00421 Thioredoxin family signature T IPR Thioredoxin Molecular Function:protein disulfide oxidoreductase activity (GO: ), Biological Process:glycerol ether metabolic process (GO: ), Biological Process:cell redox homeostasis (GO: ), Molecular Function:electron carrier activity (GO: ) A2YIW7 f927b0d241297dcc9a1c5990b58bf3c4 122 PRINTS PR00421 Thioredoxin family signature T IPR Thioredoxin Molecular Function:protein disulfide oxidoreductase activity (GO: ), Biological Process:glycerol ether metabolic process (GO: ), Biological Process:cell redox homeostasis (GO: ), Molecular Function:electron carrier activity (GO: ) A2YIW7 f927b0d241297dcc9a1c5990b58bf3c4 122 PRINTS PR00421 Thioredoxin family signature T IPR Thioredoxin Molecular Function:protein disulfide oxidoreductase activity (GO: ), Biological Process:glycerol ether metabolic process (GO: ), Biological Process:cell redox homeostasis (GO: ), Molecular Function:electron carrier activity (GO: ) Default tab-separated values output

30 ./interproscan.sh i test_proteins.fasta o test_proteins.xml goterms F xml <?xml version="1.0" encoding="utf 8" standalone="yes"?> <protein matches xmlns=" <protein> <sequence md5="f927b0d241297dcc9a1c5990b58bf3c4">maaeegvviachnkdefdaqmtkakeagkvviidftaswcgpcrfiapvfaeyakkfpgavflkvdvdelkev AEKYNVEAMPTFLFIKDGAEADKVVGARKDDLQNTIVKHVGATAASASA</sequence> <xref id="a2yiw7"/> <matches> <fingerprints match graphscan="iii" evalue=" e 7"> <signature name="thioredoxin" desc="thioredoxin family signature" ac="pr00421"> <models> <model name="thioredoxin" desc="thioredoxin family signature" ac="pr00421"/> </models> <signature library release version="41.1" library="prints"/> </signature> <locations> <fingerprints location score="0.0" pvalue="0.0" motifnumber="3" end="48" start="39"/> <fingerprints location score="0.0" pvalue="0.0" motifnumber="2" end="89" start="78"/> <fingerprints location score="0.0" pvalue="0.0" motifnumber="1" end="39" start="31"/> </locations> </fingerprints match> <hmmer2 match score="100.5" evalue=" INF"> <signature name="thioredoxin" ac="pirsf000077"> <models> <model name="thioredoxin" ac="pirsf000077"/> </models> <signature library release version="2.74" library="pirsf"/> </signature> <locations> <hmmer2 location hmm length="0" hmm end="108" hmm start="1" evalue=" e 27" score="0.0" end="113" start="4"/> </locations> </hmmer2 match>...etc XML output

31 Pre-calculated match lookup BerkeleyDB-backed REST web service Includes matches for all of UniParc (27 million sequences) 250 million matches Fast response Integrated into i5.

32 Other functionality Increased reliability Precalculated match lookup Configuration simple properties file Nucleotide sequence getorf map matches to nucleotide coordinates Pathway mapping KEGG, Reactome, MetaCyc, Unipathway

33 Future functionality Webservice Interact directly with architecture: LAN LSF PBS Sun Grid Engine Database persistence Oracle MySQL Postgres etc Graphical output Other functionality ask!

34 InterProScan 5 timeline Beta release August 2011 InterProScan 4 still maintained Full release Early 2012 InterProScan 4 deprecated interproscan-5-dev@googlegroups.com

35 Acknowledgements Team leader Developers Bioinformaticians Curators Sarah Hunter Matthew Fraser Anthony Quinn Phil Jones Craig McAnulla Alex Mitchell Sebastien Pesseat Maxim Scheremetjew Siew-Yit Yong Amaia Sangrador Any Questions Stand 302

36 Come and see us at booths 9 and 10! Job opportunities PhD and postdoc positions Training in person and online Services Industry programme EBI is an Outstation of the European Molecular Biology Laboratory.